Data Mining with WEKA(Mining Association Rules)

Explorer: finding associations

§ WEKA contains an implementation of the Apriori algorithm for learning association rules– Works only with discrete data

§ Can identify statistical dependencies between groups of attributes:– milk, butter Þ bread, eggs (with confidence 0.9 and

support 2000)

§ Apriori can compute all rules that have a given minimum support and exceed a given confidence

Data Mining with WEKA(Knowledge Flow GUI)

The Knowledge Flow GUI

§ Graphical user interface for WEKA§ Java-Beans-based interface for setting up and

running machine learning experiments§ Data sources, classifiers, etc. are beans and can

be connected graphically§ Data “flows” through components: e.g.,

“data source” -> “filter” -> “classifier” -> “evaluator”

§ Layouts can be saved and loaded again later